pytorch-tutorial/src/tensor_initialization.py at master · kozer/pytorch-tutorial · GitHub

History

67 lines (56 loc) · 1.55 KB

Raw

import torch

# ----------------------------------------- #

# Tensor initialization #

# ----------------------------------------- #

my_tensor = torch.tensor([[1, 2, 3], [4, 5, 6]])

my_tensor2 = torch.tensor([[1, 2, 3], [4, 5, 6]], dtype=torch.float32)

my_tensor3 = torch.tensor([[1, 2, 3], [4, 5, 6]],

dtype=torch.float32,

device="cuda")

my_tensor4 = torch.tensor([[1, 2, 3], [4, 5, 6]],

dtype=torch.float32,

device="cpu")

my_tensor5 = torch.tensor([[1, 2, 3], [4, 5, 6]],

dtype=torch.float32,

device="cuda", requires_grad=True)

print(my_tensor)

print(my_tensor2)

print(my_tensor3)

print(my_tensor4)

print(my_tensor5)

# Other initialization methods

x = torch.empty(size=(3,3))

print(x)

x = torch.zeros((3,3))

print(x)

x = torch.rand((3,3))

print(x)

x = torch.ones((3,3))

print(x)

x = torch.eye(5,5)

print(x)

x = torch.arange(start=0, end=5, step=1)

print(x)

x = torch.linspace(start=0.1, end=1, steps=10)

print(x)

x = torch.empty(size=(1,5)).normal_(mean=0, std=1)

print(x)

x = torch.empty(size=(1,5)).uniform_(0,1)

print(x)

x = torch.diag(torch.ones(3))

print(x)

# How to initialise and convert tensors to other types (int, float, double)

x = torch.arange(4)

print(x)

print(x.bool())

print(x.short())

print(x.long())

print(x.half())

print(x.float())

print(x.double())

#Array to tensor conversion

import numpy as np

np_arr = np.zeros((5,5))

tensor = torch.from_numpy(np_arr)

print(x)

print(tensor.numpy())

pytorch-tutorial/src/tensor_initialization.py at master · kozer/pytorch-tutorial · GitHub

Expand file tree

Breadcrumbs

tensor_initialization.py

Latest commit

History

Breadcrumbs

tensor_initialization.py

File metadata and controls